Meteo WebApp

This project is a Python/Flask web application that shows an interactive global map of meteorological stations backed by a local PostgreSQL database. Data is currently sourced from ECA&D (European Climate Assessment & Dataset). For each station, the app can render static Bokeh graphs of maximum, minimum and mean temperatures to visualize their evolution over time.

The front-end uses OpenLayers to display an OpenStreetMap base layer and clickable station markers. A toolbar allows you to filter stations by provider and country, and to switch the user interface language between English and Spanish.

Features

• Flask-based web application with an app factory (flaskr/app.py).
• PostgreSQL database for stations, providers, magnitudes, measurements and ECAD sources.
• ECAD provider integration (download, parse, and persist source files).
• Per-station temperature graphs generated as static HTML using Bokeh.
• OpenLayers map with station markers and popups.
• Toolbar with:
  • Provider filter.
  • Country filter (names localized via country-list).
  • Language selector (English/Spanish) for UI labels.

Project structure (high level)

• flaskr/app.py – Flask app factory, logging configuration, route registration and data-processing bootstrap.
• flaskr/config/config.py – Central configuration (paths, DB credentials, log file location).
• flaskr/db/ – Database connection (db.py), SQL helper methods (statements.py), schema (schema.sql).
• flaskr/data/ – Data orchestration, including ECAD provider implementation under flaskr/data/ecad/.
• flaskr/graphs/ – Bokeh utilities and ECAD-specific graph generation.
• flaskr/osmap/os_map.py – Blueprint that renders the OpenLayers-based map and toolbar.
• flaskr/templates/index.html – Main HTML template for the map UI.
• flaskr/static/ – OpenLayers library, CSS and JS assets.

Requirements

This project does not ship a requirements.txt, but you can infer the main dependencies from the imports:

• Python 3.x
• Flask
• psycopg2 or psycopg2-binary
• pandas
• bokeh
• tqdm
• country-list

Example setup with a virtual environment:

python3 -m venv .venv
source .venv/bin/activate
pip install Flask psycopg2-binary pandas bokeh tqdm country-list

Database setup

The schema targets PostgreSQL. Defaults are:

• Database name: meteo
• Host: localhost
• User: meteo
• Password: meteo

1. Create database and user (run as a PostgreSQL superuser):

sudo -u postgres psql -c "CREATE USER meteo WITH ENCRYPTED PASSWORD 'meteo';"
sudo -u postgres psql -c "CREATE DATABASE meteo OWNER meteo;"
sudo -u postgres psql -c "GRANT ALL PRIVILEGES ON DATABASE meteo TO meteo;"

2. Initialize schema and seed metadata using the Flask CLI command defined in flaskr/db/db.py:

export PYTHONPATH=$(pwd)/flaskr
flask --app flaskr/app init-db

Warning: init-db drops and recreates all tables defined in flaskr/db/schema.sql. Do not run it against a database with data you want to keep.

Running the development server

Use the Flask CLI (recommended for debug and auto-reload):

export PYTHONPATH=$(pwd)/flaskr
flask --app flaskr/app run --debug

Alternatively, run the module directly:

python3 -m flaskr.app

On startup (outside of the init-db command), the app will:

• Initialize provider definitions from the database.
• Download and process ECAD data if needed.
• Populate the database with stations and source metadata.
• Generate per-station Bokeh HTML graphs under flaskr/graphs/current/<provider>/.

This data pipeline can take a while on the first run.

Using the web UI

1. Open your browser at http://127.0.0.1:5000/.
2. Use the toolbar at the top to:
   • Select a language (English/Spanish) for the UI labels.
   • Select a provider (e.g. ecad).
   • Select a country (only countries with stations from the database are listed).
3. Click Show Map to display the stations that match the filters.
4. Hover over a marker to see a short tooltip for the station.
5. Click a marker to open a popup with more detailed information and a link to the station-temperature graph.

Useful checks and endpoints

• List map markers:

curl http://127.0.0.1:5000/stationsmarkers

• Fetch a station popup (replace 1 with a valid station ID from the stations table):

curl http://127.0.0.1:5000/popup/1

Type checking and linting

If you have mypy and pyright installed, you can run:

mypy .
pyright

Notes

• Logs are written to flaskr/logs/app.log.
• The init-db command uses flaskr/db/schema.sql as the source of truth for the database schema.
• The tests/ directory is present but does not yet contain test files; automated testing is a good next step for further development.